Documente Academic
Documente Profesional
Documente Cultură
(
m
)
WindVelocity(m/s)
Figure 3.2: Velocity profile at a distance of 89 cm from the wind tunnel outlet.
0.30
0.20
0.10
0.00
0.10
0.20
0.30
0.00 2.00 4.00 6.00 8.00 10.00
D
i
s
t
a
n
c
e
(
m
)
Windvelocity(m/s)
Figure 3.3: Velocity profile at a distance of 300 cm from the wind tunnel outlet.
Wind velocity was measured by traversing an anemometer vertically with every 50 mm
vertical position for four different locations in front of the wind tunnel exit. Figures 3.2 and 3.3
show the velocity profiles at a distance of 89 cm and 300 cm from the outlet of the wind tunnel.
18
Three Bladed Savonius rotor Models
To observe the effect of overlap ratio (ratio between the distance of the two adjacent
blades and rotor diameter) and Reynolds number on the aerodynamic characteristics of the
Savonius rotor, three different rotor models with and without overlap ratio were designed and
physically fabricated. These three models were tested in front of the subsonic wind tunnel for
various Reynolds number flow conditions.
Savonius Rotor Model 1
The three bladed Savonius rotor model called Model 1 with no overlap between adjacent
blades was designed and fabricated. Top, front, 3D design views and the fabricated scale model
of the Savonius wind turbine are shown in Figure 3.4 (a), 3.4(b), 3.4(c) and 3.4(d) respectively.
The model was made of three semi-cylindrical blades of diameter, d = 127 mm, and height, H =
300 mm. The turbine model was made of acrylic. The central shaft was removed from the turbine
model. The blades were 120 apart from each other and the overall rotor diameter was D = 248
mm for the Model 1.
19
Figure 3.4(a): Top view of Model 1 Figure 3.4(b): Front view of Model 1
Figure 3.4(c): 3D view of Model 1 Figure 3.4(d): Fabricated Model 1
20
Savonius Rotor Model 2
Savonius rotor Model 2 with overlap distance between adjacent blades, a = 25 mm, was
designed and fabricated. Top, front, 3D design views and the fabricated scale Model 2 of the
Savonius wind turbine are shown in Figure 3.5 (a), 3.5(b), 3.5(c) and 3.5(d) respectively. The
model was made of three semi-cylindrical blades of diameter, d = 127 mm, and height, H = 300
mm. The turbine model was made of acrylic without any central shaft. The blades were 120
apart from each other and the overall rotor diameter was D = 216 mm for the Model 2. Overlap
ratio for Model 2 was 0.12.
Figure 3.5(a): Top view of Model 2 Figure 3.5(b): Front view of Model 2
21
Figure 3.5(c): 3D view of Model 2 Figure 3.5(d): Fabricated Model 2
Savonius Rotor Model 3
Savonius rotor Model 3 with overlap distance between adjacent blades, a = 50 mm, was
designed and fabricated. Top, front, 3D design views and the fabricated scale Model 2 of the
Savonius wind turbine are shown in Figure 3.6 (a), 3.6(b), 3.6(c) and 3.6(d) respectively. The
model was made of three semi-cylindrical blades of diameter, d = 127 mm, and height, H = 300
mm. The turbine model was made of acrylic without any central shaft. The blades were 120
apart from each other and the overall rotor diameter was D = 192 mm for the Model 3. Overlap
ratio for Model 3 was 0.26.
22
Figure 3.6(a): Top view of Model 3 Figure 3.6(b): Front view of Model 3
Figure 3.6(c): 3D view of Model 3 Figure 3.6(d): Fabricated Model 3
23
Mathematical Expressions
Rotor Area: A = . E [3.1]
Overlap Ratio: 0R =
u
[3.2]
Aspect Ratio: AR =
H
d
[3.3]
Angular Velocity: =
2nN
60
[3.4]
Reynolds Number:Rc =
v
v
[3.5]
Tip Speed Ratio: z =
o
2v
[3.6]
Torque Coefficient: C
q
=
1
1
4
pAv
2
[3.7]
Power Coefficient: C
p
=
P
1
2
pAv
3
=
1o
1
2
pAv
3
[3.8]
Normal Drag Force: F
n
= ] p
d
2
n
0
cosJ = p
d
2
17
=1
cos
[3.9]
Tangential Drag Force: F
t
= ] p
d
2
n
0
sinJ = p
d
2
17
=1
sin
[3.10]
Normal Drag Coefficient: C
n
=
P
n
1
2
pv
2
A
[3.11]
24
Tangential Drag Coefficient: C
t
=
P
t
1
2
pv
2
A
[3.12]
Lift Coefficient: C
I
=
P
L
1
2
pv
2
A
[3.13]
Experimental Procedure
Drag Force Measurement
Figure 3.7: A three bladed Savonius rotor Model for pressure measurement.
The pressure distribution around the concave and convex surfaces of each blade were
measured experimentally using a semi-cylindrical three bladed Savonius VAWT model with
overlap distance, a = 25, between the adjacent blades as shown in Figure 3.7. The rotor model
was made of stainless steel material with each blade diameter, d = 125 mm, height, H = 300 mm
25
and rotor diameter, D =225 mm. The overlap ratio OR was 0.11 (i.e., OR = a/D) and there is no
shaft in the rotor model. The whole rotor was mounted on an iron frame using two separate
shafts and bearings at the two ends. The convex and concave surface pressures of each blade
were measured at 17 tapping points using 1.5 mm outer diameter and 10 mm length copper tubes
which were press fitted to 17 tapping holes. These tapping points were located at the mid-plane
of each blade to measure the pressure at every 10 interval on the blade surface. The copper
tubes were connected to 17 pressure transducers PX277 through 2 mm PVC tubes. Pressures
were measured at every 30 interval of rotor angle. A personal computer equipped with data
acquisition system was used to record and edit the pressure data. Average wind speed during this
experiment was 9.61 m/s. Reynolds number based on rotor diameter was 1.4710
5
. The effect of
temperature was considered in this measurement technique which was carried out at atmospheric
temperature. The normal and tangential drag forces on each blade of the Savonius rotor model
were calculated using the measured pressure difference between the concave and convex
surfaces of the blades using equations 3.9 and 3.10. Figure 3.9 shows the cross-section of the
rotor with the normal and tangential drag forces direction. Next, tangential and normal drag
coefficient was calculated using equations 3.11 and 3.12.
Static Torque Measurement
The experiment was carried out at three different wind speeds V = 9.66 m/s, 8.23 m/s and
7.33 m/s,. The Reynolds numbers based on the rotor diameter varied from 9.9410
4
to 1.610
5
.
Experiments were carried out and data was recorded at room temperature. Static torque (T) for
the three different models of the Savonius wind turbine was measured using a static torque meter
(Model: TQ-8800) at three different wind speeds. Torque meter output was in LB-inch which
26
was then converted into N-m. Rotational speed (N) was measured using a non contact photo
tachometer. Equation 3.4 was used to calculate the angular velocity from the rotational speed.
Figure 3.8: Experimental setup of wind tunnel and Savonius rotor model for static torque
measurement.
Savonius wind turbine is drag type VAWT where the lift forces are considered to be negligible.
Figure 3.8 shows the experimental setup for torque measurement of a Savonius wind turbine
model. When the wind strikes the blade surfaces of the model, two components of drag force are
generated on each blade surface. Normal drag force (F
N
) acts perpendicular on the blade surface
and tangential drag force (F
T
) acts along tangential direction on each blade. Figure 3.9 shows the
schematic diagram of the Savonius rotor cross-section with the components of drag forces on
each blade. The pressure difference between the concave and convex surfaces on each blade
produces these tangential and normal drag forces. These components of drag forces are
27
responsible for torque generation within the turbine shaft which can be measured using a torque
meter. Equation 3.7 is used to calculate the torque coefficient from the measured torque value.
Power coefficient can be calculated from the measured torque and angular velocity of the rotor
Figure 3.9: Schema
using equation 3.9.
tic diagram of the rotor model cross-section showing the normal and
tangential drag forces on each blade.
28
Selection of Numerical Model
cal model from the Computational Fluid Dynamics (CFD)
code FLUENT 2D, a NACA 4412 airfoil was numerically examined at the different angle of
attack and compared with established research results. For 2D modeling a mesh was generated
around the airfoil using GAMBIT. Total of 12,451 nodes were generated and clustering was
imposed near the boundary of the airfoil where the mesh was relatively coarser. Figure 3.10
shows the mesh around the airfoil after exported to FLUENT.
Figure 3.10: Mesh around the NACA 4412 airfoil using FLUENT.
The numerical d flow
condition as was used by the previous research for verification. These conditions are wind speed,
V = 45.48 m/s, Reynolds number, Re = 3,000,000, air density, r = 1.225 kg/m , air viscosity,
= 1.857110
-5
, operating pressure P = 101325 Pa. Investigation was performed for three different
angles of attack 2,4and 8.Investigation was carried out using two different numerical models
For the selection of the numeri
testing of the airfoil was conducted using the same ambient and flui
3
29
such as inviscid model and standard k- two equation turbulence model with the enhance wall
treatment options. Inviscid flow is suitable for higher Reynolds number where the effect of
viscosity is negligible and where initial force is dominating over viscous force. So inviscid flow
is very appropriate for aerodynamic analysis where this type of analysis will give a quick
estimation of drag and lift forces action on the body surface [6]. Since the k- turbulence model
was proposed by Launder and Spalding in 1972 [16], it became popular for its robustness and
accuracy for a wide range of turbulence flow. The standard k- model is a semi-empirical model
which is based on model transport equations for turbulence kinetic energy k and dissipation rate
[6].
Figure 3.11: Pressure Contours around the airfoil model at 2 angle of attack (Inviscid flow)
Figures 3.11 show the pressure contours generated at 2 angle of attack using the inviscid
model. Low pressure zone was established at the middle of the airfoil for the angle of attack 2
then it propagated towards the leading edge of the airfoil with the increase of the angle of attack.
Similar pattern was observed for the velocity counters where the higher velocity zone propagated
30
towards the leading edge with the increase of the angle of attack. Figures 3.12 and 3.16 show
velocity contours and velocity vectors for Inviscid flow at 2 angle of attack.
Figure 3.12: Velocity Contours around the airfoil model at 2 angle of attack (Inviscid flow)
Figure 3.13: Velocity vectors around the airfoil model at 2 angle of attack (Inviscid flow)
31
Using the flow simulation results from both the inviscid model and the k- turbulence
model lift coefficient was calculated using equation 3.13 and then compared with published [25]
NACA 4412 airfoil results. Table 3.1 shows the comparison of the lift coefficient values
calculated using those two models and previous research. Figure 3.14 also shows the graphical
comparison of the lift coefficient values from these three different sources. After comparing the
inviscid model and the k- turbulence model results with established published results it was
found that the k- turbulence model gave more accurate results than the inviscid model.
Therefore, the k- turbulence model has been selected for the numerical modeling of the
Savonius wind turbine.
Table 3.1:
Lift Coefficient Comparison
Angle of
Attack ()
Lift coefficient
(Inviscid Model)
Lift coefficient
(k- Model )
Lift coefficient
(Report 563)[25]
2 0.788 0.678 0.501
4 0.990 0.846 0.667
8 1.404 1.156 1.024
32
0.000
0.500
1.000
1.500
2.000
2 4 8
L
i
f
t
C
o
e
f
f
i
c
i
e
n
t
AngleofAttack
Inviscid
kModel
Report563
Figure 3.14: Comparison of NACA 4412 airfoil lift coefficient using two numerical models and
the previous established model
Numerical Procedure
The k- turbulence model was used for the computational flow simulation around the
Savonius rotor models with different overlap ratios. Commercially available software FLUENT
was used to solve the turbulent flow field and GAMBIT was used for mesh generation around
the rotor models. Numerical simulation provides the pressure and velocity values at all nodal
points of flow domain around the rotating blades. The purpose of this analysis was to observe the
performance of various Savonius VAWTs configurations; it was felt that a 2-D simulation was
sufficient for this application. Figures 3.15, 3.16, 3.17 show 2-D mesh generated within a
computational domain around three bladed Savonius wind turbine models using GAMBIT in
which the position of the three blades were 0, 120 and 240. The size of the computational
domain was 1.6 m 1.4 m and the total number of nodes was around 39992. A size function was
introduced with the rotor blade to get the better computational results adjacent to the blade
33
surface. To introduce this size function the minimum mesh size near the blade surface was
chosen 0.0005 m 0.0005 m with the growth rate of 1.1 and maximum mesh size was 0.008 m
0.008m. In this study triangular mesh was chosen over quadrilateral mesh to reduce the
computational time. These generated meshes were then exported in FLUENT for post
processing. The flow of air within the domain around the rotor model was assumed to be
turbulent and the effects of molecular viscosity were considered negligible. The simplest
"complete models'' of turbulence are two-equation models in which the solution of two separate
transport equations allows the turbulent velocity and length scales to be independently
determined.
Figure 3.15: Generated mesh using Gambit for Model 1.
34
Figure 3.16: Generated mesh using Gambit for Model 2.
ysis of turbulent
flow around rotor models. The pressure-velocity coupling is achieved using the well-known
SIMPLE (Semi-Implicit Method for Pressure-Linked Equations) method by Patankar [22].
Turbulent kinetic energy (k) and turbulent dissipation rate () first order upwind scheme was
chosen for the momentum equation solution. The standard k- turbulence model [16] is a semi-
Figure 3.17: Generated mesh using Gambit for Model 3.
The standard k - turbulence model in FLUENT was used for the anal
35
empirical model based on model transport equations for the turbulence kinetic energy (k) and its
dissipation rate (). The model transport equation for k was derived from the exact equation,
while the model transport equation for () was obtained using physical reasoning and bears little
resemblance to its mathematically exact counterpart.
The turbulence kinetic energy, k, and its rate of dissipation, , were obtained from the
following transport equations:
t
(pk) +
x
i
(pku
) =
x
]
_[p +
t
c
k
k
x
]
_ +0
k
+0
h
- pe -
M
+S
k
[3.14]
t
(pe) +
x
i
(peu
) =
x
]
_[p +
t
c
e
e
x
]
_ +c
1e
e
k
(0
k
+C
3e
0
b
) -C
2e
p
e
2
k
+S
e
[3.15]
In these equations, G
k
represents the generation of turbulence kinetic energy due to the
mean velocity gradients, G
b
is the generation of turbulence kinetic energy due to buoyancy, Y
M
represe
C
k
s
nts the contribution of the fluctuating dilatation in compressible turbulence to the overall
dissipation rate, C
1
, C
2
and C
3
are constants.
K
and
is a constant.
igned top and
with inlet free stream ve
36
m/s, air
m/s.
density was considered 1.2 kg/m
3
. The blades were considered as moving walls and their
rotational velocity was provided from the rpm measured during the experiment. The convergence
of the sequential iterative solution is achieved when the sum of the absolute differences of the
solution variables between two successive iterations falls below a pre-specified small number,
which was chosen as 110
5
in this study. For all Models using k- turbulence model
convergence criteria(110
5
) was set and tested for continuity, x-velocity, y-velocity, kinetic
energy (k) and turbulent dissipation rate (). It was found that for Model 1 at wind speed 9.66
m/s solution converged at 4019 iterations shown in Figure 3.18. For Model 2 and Model 3 at the
wind speed 9.66 m/s solution converged respectively at 1844 and 1307 iterations.
Figure 3.18: Residuals Convergence of Model 1 at wind speed 9.66
37
CHAPTER 4
FINDINGS OF THE STUDY
Introduction
Experimental results of normal drag coefficient, tangential drag coefficient, torque
coefficient, and power coefficient of three bladed Savonius VAWTs models with various overlap
ratios are discussed in this chapter. The second part of this chapter will discuss the numerical
findings of the present study as well as the comparison of torque and power coefficients with the
experimental results. Both numerical and experimental investigations were carried out at
different wind speed for different Reynolds number.
Experimental Results
Normal and Tangential Drag Coefficient
Normal drag coefficient (C
n
) variation with the change in rotor angle () for three bladed
Savonius wind turbine model is shown in Figure 4.1. Combined blade effect at 10 interval
from 0 to 360 is shown in this plot. Normal drag coefficient increases with the increase of
rotor angle from 0 to 60, and then decreases with the increase of rotor angle up to 100.
Normal drag coefficient is responsible for torque generation in the rotor model. Figure 4.1 also
shows that this combined normal drag coefficient remains positive for rotor angle changing
from 0 to 90 with its maximum value at 60 which is favorable for torque production. The
same pattern of normal drag coefficient repeats from 120 to 230 and from 240 to 350.
38
Tangential drag coefficient (C
t
r angle () with 10 interval from 0
to 360 is shown in Figure 4.2. Figure also shows that a sharp drop occurs from 0 to 10 and
then sharp increase occurs urs in the drag coefficient
from 40 to 90 and a sharp increase occurs from 90 to 120. Results show that tangential drag
ins positive with every angle of rotation this positive value is very important
factor for producing thrust in the rotor model. The same pattern of tangential drag coefficient
repeats from 120 to 230 and from 240 to 350.
) with change in roto
from 10 to 40. Again a sharp drop occ
coefficient rema
Figure 4.1: Normal Drag Coefficient (C
n
) versus Angle of rotation () for three
blades combined effect.
0.4
0.2
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
0
0.4
0.6
2
0
0
2
2
0
2
4
0
2
6
0
2
8
0
3
0
0
3
2
0
3
4
0
3
6
0
N
o
r
m
a
l
n
)
0.8
1
1.2
C
o
e
f
f
i
c
i
e
n
t
(
C
0.2
D
r
a
g
Angleofrotation()
39
0
0.2
0.4
1
1.2
1.4
0
D
r
a
g
C
o
e
f
f
i
c
i
e
n
t
(
C
t
)
0.6
0.8
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
2
0
0
2
2
0
2
4
0
2
6
0
2
8
0
3
0
0
3
2
0
3
4
0
3
6
0
T
a
n
g
e
n
t
i
a
l
Angleofrotation()
Figure 4.2: Tangential Drag Coefficient (C
t
) versus Angle of Rotation () for
three blades combined effect.
Torque Coefficient
Torque Coefficient Variation for Three Individual Savonius VAWT Models:
Torque coefficient of Savonius wind turbine Model 1 was calculated for three different
Reynolds num
of rotation (). Torque coefficient was calculated for combined blade effect at every 30 interval
0 to 360. Three Reynolds numbers for Model 1 were 1.6110
5
(for wind speed 9.66
m/s), 1.3710
5
(for wind speed 8.23 m/s), and 1.2210
5
(for wind speed 7.33 m/s). For every
Reynolds number the values of torque coefficient increase from 0 to 60 and then start to
decrease from 60 to 120. The same pattern repeats for the blade angle from 120 to 210 and
from 240 to 330. For Reynolds number 1.0010
5
, the value of torque coefficient is negative at
ber. Figure 4.3 shows torque coefficient (C
q
) variation with the increase of angle
from
40
120, 210 and 240. It is desired to remove the negative torque for all rotor position , as this
negative torque causes reverse rotation which can reduce power output. Figure 4.4 shows
torque coefficient (C
q
) variation with the angle of rotation () for Model 2. Similarly combined
blade effect on torque coefficient with 30 interval from 0 to 360 was calculated. Reynolds
numbers for Model 2 were 1.4010
5
(for wind speed 9.66 m/s), 1.1910
5
(for wind speed 8.23
m/s), and 1.0610
5
(for wind speed 7.33 m/s). From the figure it can be seen that the torque
coefficient increases from 0 to 60 and decreases at 90 again increases at 120 (except at Re =
1.0610
5
). There was no negative torque coefficient for this model. The same pattern repeats
for the blade angel from 120 to 210 and from 240 to 330. Figure 4.5 shows torque
coefficient (C ther
two models, c e coefficient at every 30 interval from 0 to 360
bers for Model 3 was 1.2410
5
(for wind speed 9.66 m/s),
1.0610
5
(for wind speed 8.23 m/s) and 9.4410
5
(for wind speed 7.33 m/s. For Re = 1.2410
5
,
e pattern of the graph looks similar. Torque coefficient increases from
0
5
) . The same
pattern repeats for the blade angel from 120 to 210 and from 240 to 330.
q
) variation with the increase of angle of rotation () for Model 3. Likewise o
ombined blade effect on torqu
was calculated. Reynolds num
1.0610
5
and 9.4410
5
th
0 to 60 then decrease at 90 and again increase at 120 (except for Re = 1.241
41
Figure 4.3: Torque Coefficient (C
q
) versus Angle of Rotation () for Model 1.
Figure 4.4: Torque Coefficient (C
q
) versus Angle of Rotation () for Model 2.
0.300
0.200
0.100
0.100
0.200
0.400
0.500
0.700
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
T
o
r
q
u
e
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
0.000
0.300
0.600
AngleofRotation()
Re=1.61E+05
Re=1.37E+05
Re=1.22E+05
0.000
0.100
0.200
0.300
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
T
o
r
q
u
e
AngleofRotation()
0.400
0.500
0.600
0.700
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
Re=1.40E+05
Re=1.19E+05
Re=1.06E+05
42
Figure 4.5: Torque Coefficient (C
q
) versus Angle of Rotation () for Model 3.
Torque Coefficient Variation at Three Different Wind Speeds:
0.000
0.100
0.200
0.300
0.400
0.500
0.600
0.700
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
T
o
r
q
u
e
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
AngleofRotation()
Re=1.24E+05
Re=1.06E+05
Re=9.44E+04
Torque coefficient (C
q
) variation with the angle of rotation () for three models at a wind
speed of 9.66 m/s is shown in Figure 4.6. Reynolds numbers 1.6110
5
, 1.4010
5
, and 1.2410
5
were used for three models respectively with a 30 interval from 0 to 360. Patterns of the graph
for all three models is similar; torque coefficient increases from 0 to 60 and decreases from 60
to 120 (except for model 2). The similar patterns repeat from 120 to 210 and from 240 to
330. Model 1 exhibits negative torque coefficient at 120 and 240. Figure 4.7 shows torque
coefficient variation with the angle of rotation for three different models at wind speed of 8.23
m/s. Reynolds numbers 1.3710
5
, 1.1910
5
, and 1.0610
5
were used for Model 1, Model 2 and
Model 3 respectively. Figure 4.7 also shows that torque coefficient increases from 0 to 60 and
then decreases from 90 to 120 (except for model 1which keeps on increasing till 90). The
same patterns repeat from 120 to 230 and from 240 to 330. Negative torque coefficient is
43
observed at 120, 210 and 240 in Model 1. Figures 4.8 show the torque coefficient variation
with the angle of rotation for three different models at wind speed of 7.33 m/s. The pattern of the
plot is similar to the previous plots. From these three figures it can be concluded that with the
increase of wind speed the corresponding values of torque coefficients decrease as the wind
speed increases.
Figure 4.6: Torque Coefficient (C
q
) versus Angle of Rotation () at wind speed of 9.66 m/s.
0.200
0.000
0.100
0
0.300
0.400
0
3
0
6
0
9
0
3
6
0
T
o
r
q
u
e
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
.200
Model1
Model2
Model3
0.100
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
AngleofRotation()
44
0.400
0.200
0.000
0.200
0.400
0.600
0.800
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
T
o
r
q
u
e
AngleofRotation()
Model1
Model2
Model3
Figure 4.7: Torque Coefficient (C
q
) versus Angle of Rotation () at wind speed of 8.23 m/s.
Figure 4.8: Torque Coefficient (C
q
) versus Angle of Rotation () at wind speed of 7.33 m/s.
Comparisons of Experimental Torque Coefficient with Previous Research Results:
0.000
Figure 4.9 shows a comparison of experimental torque coefficient (C
q
) variation of three
models with the previous experimental results by Hayashi et al. [19]. They have used two bladed
0.100
0.200
T
o
r
q
u
e
0.300
0.400
0.500
0.600
0.700
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
AngleofRotation()
Model1
Model2
Model3
45
0.200
0.100
0.000
0.100
0.200
0.300
0.400
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
T
o
r
q
u
e
C
o
e
f
f
i
c
i
e
n
t
(
C
q
)
AngleofRotation()
Model1
Model2
Model3
HayashietalModel
Savonius wind turbine model with blade diameter, d =184 mm, overall rotor diameter, D = 330
mm, height, H = 230 mm and overlap ratio for the model was 0.2. They carried out the
experiment at a wind speed of 9 m/s. Reynolds number used for their model was 2.1010
5
.
For
the current Model 1, Model 2 and Model 3 experiment was carried out at wind speed of 9.66 m/s
and the corresponding Reynolds numbers were 1.6110
5
, 1.40 10
5
, and 1.2410
5
respectively.
Currently the Reynolds number used in Model 1 is close to the Reynolds number of Hayashi et
al. Models and the two graphs show similar pattern for torque coefficient variation only
difference is that for Hayashi Model torque coefficient variation graph shifted towards right than
current Model 1. This difference is due to the fact that their model was two-bladed rotor whereas
ient
values are in very close match with the current study of Model 1.
Figure 4.9: Comparison of Torque Coefficient (C
q
) versus Angle of Rotation ().
the current models all are three-bladed rotors. The maximum and minimum torque coeffic
46
Power Coefficient
Power Coefficient Variation for Three Individual Savonius VAWT Models:
icient (C
P
) variation with the
change of angle of rotation () from 0 to 360 for Model 3. For this model, power coefficient
variation follows the similar trend for all Reynolds number, increasing from 0 to 60 and then
decreasing up to 120 then repeats from 120 to 230and from 240 to 330.
Figure 4.10: Power Coefficient (C
P
) versus Angle of Rotation () for Model 1.
Power coefficient (C
P
) was calculated at three different Reynolds number (Re) for all
three models. Figure 4.10 shows power coefficient (C
P
) variation with angle of rotation ()
from 0 to 360 for Model 1. Trends of the plots are similar for Re = 1.6110
5
, 1.3710
5
and
1.2210
5
. Power coefficient was negative at 120, 210 and 240 for Re = 1.3710
5
and at 120
and 240 for Re = 1.6110
5
. For this model, better power coefficient variation occurred at Re =
1.2210
5
. Figure 4.11 shows power coefficient (C
P
) variation with angle of rotation () at three
different Reynolds number (Re) for Model 2. There is no negative power coefficient for this
model at any Reynolds number. Figure 4.12 shows power coeff
0.100
0.050
0.000
0.050
0.100
0.150
0.200
0.250
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
p
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
AngleofRotation()
Re=1.61E+05
Re=1.37E+05
Re=1.22E+05
47
Figure 4.12: Power Coefficient (C
P
) versus Angle of Rotation () for Model 3.
Figure 4.11: Power Coefficient (C
P
) versus Angle of Rotation () for Model 2.
0.000
0.020
0.040
0.060
0.080
3
0
6
0
9
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
P
o
w
e
r
0.160
0.100
0.140
00
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
AngleofRotation()
0.120
Re=1.40E+05
Re=1.19E+05
Re=1.06E+05
0.060
0.000
0.010
0.020
0.030
0.040
0.050
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
AngleofRotation()
Re=1.24E+05
Re=1.06E+05
Re=9.44E+04
48
Power Coefficient Variation at Four Different Wind Speeds:
igures 4.13 to 4.15 show power coefficient (C
P
) variation with the increase of angle of
rotation (
P
variation at wind speed of 9.66 m/s where improved power coefficient is observed for Model 2
and Model 3 than Model 1. Negative power coefficient values occur at rotor angle 120 and 240
for Model 1. Figure 4.14 shows that Model 2 and Model 3 demonstrate better power coefficients
than Model 1 while wind speed is at 8.23 m/s and negative power coefficient occurs at rotor
angle 120 and 240 for Model 1. Figure 4.15 shows that for all three rotor models power
coefficient remains positive with the increase of angle of rotation at wind speed of 7.33 m/s.
Figure 4.13: Power Coefficient (C
P
) versus Angle of Rotation () at wind speed 9.66
m/s.
F
) for three Savonius rotor models at three different wind speeds. Figure 4.13 shows C
0.100
0.050
0.000
0.050
0.100
0.150
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
AngleofRotation()
(
C
p
)
Model1
Model2
Model3
49
Figure 4.14: Power Coefficient (C
P
) versus Angle of Rotation () at wind speed 8.23 m/s.
Figure 4.15: Power Coefficient (C
P
) versus Angle of Rotation () at wind speed 7.33 m/s.
0.100
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
Angleofrotation()
0.050
0.000
0.050
0.100
0
0.200
p
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
.150
Model1
Model2
Model3
0.000
0.050
0.100
0.150
0.200
0.250
0
3
0
6
0
9
0
1
2
0
1
5
0
1
8
0
2
1
0
2
4
0
2
7
0
3
0
0
3
3
0
3
6
0
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
AngleofRotation()
Model1
Model2
Model3
50
Nume
sure Contours for Three Models at Three Different Reynolds Number
ressure contours generated from numerical simulation of Model 1 for three different
Reynolds number are shown in Figures 4.16 to 4.18. For all these cases higher pressure values
were found at the convex side of the first blade Savonius rotor model. Negative pressure region
was developed from convex side of blade 2 to some portion of convex side of blade 3. This
negative pressure is creating pressure difference between concave and convex surface that
eventually rotates the turbine blades.
Figure 4.16: Pressure Contour around Savonius rotor Model 1 at Re = 1.6110
5
.
rical Results
Pres
P
51
Figure 4.18: Pressure Contour around Savonius rotor Model 1 at Re = 1.22 10
5
.
Pressure contours around the Savonius rotor Model 2 at three different Reynolds number
are shown in Figures 4.19 to 4.21. Likewise Model 1 pressure contours the higher pressure
region was developed at convex side of blade 1 and most negative pressure region was
Figure 4.17: Pressure Contours around Savonius rotor Model 1 at Re = 1.37 10
5
.
52
developed at outside of convex side of blade 3. Pressure contours around the Savonius rotor
: Pressure Contour around Savonius rotor Model 2 at Re = 1.40 10
5
.
Figure 4.20: Pressure Contour around Savonius rotor Model 2 at Re = 1.19 10
5
.
Model 3 are shown in Figures 4.22 to 4.24.
Figure 4.19
53
Figure 4.21: Pressure Contour around Savonius rotor Model 2 at Re = 1.06 10
5
.
Figure 4.22: Pressure Contour around Savonius rotor Model 3 at Re = 1.24 10
5
.
54
Figure 4.23: Pressure Contour around Savonius rotor Model 3 at Re = 1.06 10
5
.
Figure 4.24: Pressure Contour around Savonius rotor Model 3 at Re = 9.44 10
4
.
55
Velocity Contours for Three Models at Three Different Reynolds Number
Contours of Velocity magnitude for Savonius rotor Model 1 at three different Reynolds
number are shown in Figures 4.25 to 4.27. Patterns of the contours are almost same for different
Reynolds number only exception is a slight variation in velocity magnitude. Once the wind
strikes the turbine blades the velocity starts to decrease at the trailing edge of the Savonius wind
turbine model but after some distance travel stars to regain the velocity. Higher velocity region
was created at the top and bottom side of the wind turbine model. Figures 4.28 to 4.30 show
velocity contours for Model 2 and Figures 4.31 to 4.33 show velocity contours for Model 3 at
ilar patterns of velocity contours are observed for Model 2 and
Model 3 but only the velocity magnitudes are different for different model cases. From these
that with the increase of overlap ratio the lower velocity region shorten
after the trailing edge and come closer to the turbine blades.
Figure 4.25: Velocity Contour around Savonius rotor Model 1 at Re = 1.6110
5
.
different Reynolds number. Sim
figures it can be seen
56
Figure 4.26: Velocity Contour around Savonius rotor Model 1 at Re = 1.3710
5
.
Figure 4.27: Velocity Contour around Savonius rotor Model 1 at Re = 1.22 10
5
.
57
Figure 4.28: Velocity Contour around Savonius rotor Model 2 at Re = 1.40 10 .
Figure 4.29: Velocity Contour around Savonius rotor Model 2 at Re = 1.19 10
5
.
5
58
Figure 4.30: Velocity Contour around Savonius rotor Model 2 at Re = 1.06 10
5
.
Figure 4.31: Velocity Contour around Savonius rotor Model 3 at Re = 1.24 10
5
.
59
Figure 4.32: Velocity Contour around Savonius rotor Model 3 at Re = 1.06 10
5
.
Figure 4.33: Velocity Contour around Savonius rotor Model 3 at Re = 9.44 10
4
.
60
Torque Coefficient (C
q
)
Figure 4.34 shows the numerically calculated torque coefficient (C
q
) variation with
different Reynolds number (Re) for three different models. With the increase of Reynolds
number torque coefficient slightly increases for all three models. Model 1 gives better torque
coefficient compared to other two models.
Figure 4.34: Torque Coefficient (C
q
) versus Reynolds number (Re) for three Models.
Figure 4.35 shows the comparison of numerically and experimentally calculated power
coefficient (C
P
) of the three Savonius rotor Models with the increase of tip speed ratio ().
Converged solutions of the power coefficient values were considered at all tip speed ratios for
0
0.002
0.004
0.006
0.008
0.012
0
0.016
5
.
0
0
E
+
0
4
7
.
5
0
E
+
0
4
1
.
0
0
E
+
0
5
1
.
2
5
E
+
0
5
1
.
5
0
E
+
0
5
T
o
r
q
u
e
(
C
q
)
ReynoldsNumber(Re)
.014
0.01
C
o
e
f
f
i
c
i
e
n
t
Model1
Model2
Model3
Comparison of Numerical and Experimental Power Coefficient
61
62
0
0.021 0.35 0.39 0.47
0.05
0.1
0.15
0.2
0.25
0.3
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
TipSpeedratio()
NumericalCp
ExpCpat0Angle
ExpCpat30Angle
ExpCpat60Angle
ExpCpat90Angle
numerical results whereas power coefficient at four rotor positions 0, 30, 60 and 90 were
considered for experimental values. Combined blade effect was considered for both experimental
and numerical calculation. Figure 4.35 shows that for Model 1 experimental power coefficient at
rotor position 0 is very close to the numerical results. But the deviation is huge for rotor
position 60. This may be during experiment there was disturbance from environment which
causes sudden power increase and also in numerical simulation the boundary effect causes lower
numerical value. Similar pattern of power coefficient variation is observed for Model 2 and
odel 2
nd Model 3 for both numerical and experimental cases.
Figure 4.35: Power Coefficient (CP) versus Tip speed ratio () for Model 1.
Model 3 as shown in Figures 4.36 and 4.37. But the magnitude of the C
P
decreases for M
a
63
0
0.03
0.06
0.028 0.19 0.24 0.35
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
TipSpeedratio()
0.09
0.12
0.15
0.18
(
C
p
)
NumericalCp
ExpCpat0Angle
ExpCpat30Angle
ExpCpat60Angle
ExpCpat90Angle
0
0.03
0.06
0.09
0.12
0.15
0.18
0.028 0.19 0.24 0.35
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
TipSpeedratio()
NumericalCp
ExpCpat0Angle
ExpCpat30Angle
ExpCpat60Angle
ExpCpat90Angle
Figure 4.36: Power Coefficient (C
P
) versus Tip speed ratio () for Model 2.
Figure 4.37: Power Coefficient (C ) versus Tip speed ratio () for Model 3.
0
0.04
0.06
0.08
0.1
P
o
w
e
r
C
o
e
f
f
i
c
i
e
n
t
(
C
p
)
ComputationalCp
ExpCpat0Angle
ExpCpat30Angle
ExpCpat60Angle
0.02
ExpCpat90Angle
0.017 0.089 0.092 0.16
TipSpeedratio()
P
CHAPTER 5
CONCLUSION AND RECOMMENDATION
64
Introduction
Three different three bladed Savonius wind turbine scale models with different overlap ratios
(Model 1: no overlap, Model 2: overlap ratio 0.12 and Model 3: overlap ratio 0.26) were
designed and fabricated for the current study. Aerodynamic characteristics of these models were
experimentally investigated using the subsonic wind tunnel. Experimental investigation was
performed at different Reynolds numbers. Numerical investigation was also performed to
determine torque and power coefficients using GAMBIT and FLUENT.
Conclusion
From this current study, analysis and results of this research work, the following conclusions can
be made:
i. For Model 1 with Re = 1.2210
5
, Model 2 with Re = 1.0610
5
, and Model 3 with Re =
9.9410
4
experimental torque coefficient (C
t
) shows higher and positive values compared
to other Reynolds numbers. It also shows that lower Reynolds number gave better torque
coefficient (C
t
) variation with the increase of the angle of rotation for each Model. Model
2 demonstrate better experimental torque coefficient (C
t
) for all three different wind
speeds (9.66 m/s, 8.23 m/s, and 7.33 m/s).
ii. For Model 1 with Re = 1.2210
5
, Model 2 with Re = 1.1910
5
, and Model 3 with Re =
9.9410 experimental power coefficient (C
P
) shows higher and positive values
compared to other Reynolds numbers. Model 2 shows the better experimental power
4
65
coefficient (C
P
) at wind speed 9.66 m/s and wind speed 8.23 m/s. But for wind speed
7.33 m/
P
).
iii. With the increase of Reynolds number numerical torque coefficient increases for all
three Models.
iv. Power coefficient calculated from numerical method shows that it is always increasing
with the increase of tip speed ratio. For Model 1 numerical power coefficient matches
well with the corresponding experimental values at 0 rotor position.
Suggestion of future work
Following suggestions can be made for further improvement of the current research:
Present nolds
numbers from 7.7810
4
to 1.61 10
5
.
For future study it is recommended that
investigation can be perform for higher tip speed ratios for better understanding of the
change of power and torque characteristics variation with higher tip speed ratios.
Also the effect of aerodynamic characteristic can be observed at higher Reynolds
numbers. For this, a modified subsonic wind tunnel which will have the option of
variable wind speed as well as higher velocity limit is strongly recommended for better
experimental investigation.
Dynamic torque can be measured using the dynamic torque tester for better results and
thus an efficient model can be designed and improved results can be obtained.
Better aerodynamic performance can be achieved by improving the VAWT design by
optimiz
s Model 1 shows the better power coefficient (C
study was conducted for tip speed ratios from 0.017 to 0.47 and for Rey
ing the blade shape and blade numbers.
Investigation can be made by types of vertical axis wind turbine
(VAWT) like Darrious and Savonius wind turbine to overcome the negative torque
generate
combining the two
d during the rotation.
66
BIBILOGRAPHY
[1] Aldoss, T. K., Obeidat, K. M. (1987) Performance analysis of two Savonius
rotors running side by side using the discrete vortex method, Wind
nd rotor. Energy
nd management, 3425-3432.
al study on
improvement of a savonius rotor performance with curtaining.
Experimental thermal and fluid science, 32, 1673-1678
[4] Benjanirat, S., Sankar, L., & Xu, G. (2003). Proceedings of 41st aerospace
sciences meeting and exhibit Evaluation of turbulence models for the
prediction of wind turbine aerodynamics. Reno, Nevada.
[5] Diaz, F., Gavalda, J., Massons, J. (1991) Drag and lift coefficients of the
Savonius wind machine, Journal of Wind Engineering. 15, 240-246.
[6] Fluent 6.3 user guide
[7] Fujisawa , N. (1992). On torque mechanism of savonius rotors. Journal of
Wind Engineering and Industrial Aerodynamics, 40, 227-292
[8] Fujisawa, N., Gotoh, F. (1994) Experimental study on the aerodynamic
Engineering, 11, 79-88.
[2] Altan, B. D., Atilgan, M. (2008). An experimental and numerical study on the
improvement of the performance of savonius wi
conversion a
[3] Altan, B. D., Atilgan, M, Ozdamar, A. (2008). An Experiment
performance of a Savonius rotor, ASME Journal of solar energy
engineering, 116, 148 152.
67
[9] Gupta, R., Biswas, A., Sharma, K. K. (2008). Comparative study of a three
bucket savonius rotor with a combined three bucket savonius three
bladed darrieus rotor. Rnewable Energy, 33, 1974-1981
[10] Hau, E. (2006) Wind turbines: Fundamentals, technologies, applications,
economics. Germany. Springer- verlag berlin Heidelberg.
04) Proceedings of European Wind
Energy Conference and Exhibition, Wind tunnel test on a three stage out
phase Savonius rotor, London, England.
[12] Hayashi, T. , Li, Y., Hara, Y. (2005). Wind tunnel tests on a different phase
three stage savonius rotor. JSME international journal, 48, 9-16.
[13] Islam, A. K. M. S., Islam, M. Q., Razzaque, M. M.,& Ashraf, R., (1995), Static
Torque and Drag Characteristics of an S-shaped Savonius Rotor and
Prediction of Dynamic Characteristics, Wind Engineering, 19.
[14] Kamoji, M.A. , Kedare, S.B. , Prabhu, S.V. (2009). Performance test of helical
savonius rotors. Renewable Energy, 34, 521-529
[15] Kamoji, M. A., Kedare, S. B. (2007) Proceedings of 5th AIAA International
Energy Conversion Engineering Conference. Wind tunnel tests on a
single stage helical Savonius rotor.
[16] Launder, B. E. and Spalding, D. B.( 1972) Lectures in Mathematical Models of
Turbulence. Academic Press.
[11] Hayashi, T., Li, Y., Hara, Y., Suzuki (20
68
[17] Lida, A., Kato, K. & Mizuno, A.(2007). Proceedings of 16
th
Australasian fluid
mechanics conferences. Numerical simulation of unsteady flow and
[18] . (2008). Velocity measurement of flow around
[19] gs of European Wind Energy
[20] alculation of the
[21] wind energy:
) Numerical Heat Transfer and Fluid Flow, Hemisphere
[23] n Darrieus
[24]
nd Turbine Deployments: Market Analysis
and Forecasts. Boulder, CO. : Beh, G. , Wheelock, C.
aerodynamic performance of vertical axis wind turbine with LES. Gold
Coast, Australia.
McWilliam, M., Johnson, D.A
model vertical axis wind turbines. International Journal of Green
Energy,5, 55-68.
Menet, J.L., Bourabaa, N. (2004). Proceedin
Conference. Increase in the Savonius Rotor Efficiency via a Parametric
Investigation. London, UK
Menet, J.L., Valdes, L.C., Menart, B. (2001). A comparative c
wind turbines capacities on a basis of the L criterion, Renewable
Energy, 22, 491506.
Newman, B. G. (1974). Proceedings of symposium on
achievements and potential. Measurements on a savonius rotor with a
variable gap. Sherbrooke, Canada.
[22] Patankar, S. V., (1980
Publishing Corporation.
Paraschivoiu, I. (2002). Wind turbine design: with emphasis o
concept. Montreal, Canada: Polytechnic International Press.
Pike Research LLC (2009). Wind Energy Outlook for North America Wind
Power Generation Capacity a
69
[25] Pinkerton, R. B. (1936). Calculated and measured Pressure distribution over
the midspan section of the NACA 4412 airfoil. NACA report no 563.
365-380
[26] Rahman, M., Islam, M. Q., Islam, A.K.M.S. (1999) Proceeding of the third
international conference on fluid mechanics and heat transfer. Prediction
[27]
Bangladesh University of Eng.
[28] nd int.
[29] ed from
ics. 96, 1359-1375.
neural networks for estimation of the power
Simulation modeling practice and theory. 17, 1290-1298
[32] Savonius, S. J. (1931). The S rotor and its application. Mechanical
Engineering, 53, 333-338.
of dynamic characteristics of a three bladed Savonius rotor.
Rahman, M. (1999) Torque and drag characteristics of a three bladed Savonius
rotor, M. Sc. Thesis, Mech. Eng. Dept.,
and Tech., Bangladesh.
Rahman, M., Islam, M. Q., Islam, A.K.M.S. (1999) Proceeding of the 2
seminar on renewable energy for poverty alleviation, IEB, Bangladesh,
Aerodynamic characteristics of a three bladed Savonius rotor.
Riegler, H. (2003). HAWT versus VAWT, Retriev
http://www.victordanilochkin.org/research/turbine/papers/HAWT%20ver
sus.pdf
[30] Saha, U. K., Thotla, S., Maity, D. (2008). Optimum design configuration of
Savonius rotor through wind tunnel experiment. Journal of wind
engineering and industrial aerodynam
[31] Sargolzaei, J., Kianifar, A. (2009). Modeling and simulation of wind turbine
savonius rotor using artificial
ratio and torque.
70
71
[34]
pace Administration office of management.
[37] afts of Persia, their development,
[33] Sawada, T., Nahamura, M., Kamada, S. (1986) Blade force measurement and
flow visualization of Savonius rotors, Bulletin of JSME, 29, 2095-2100.
Shepherd, D. G. (1990). Historical Development of the windmill. National
Aeronautics and S
[35] U. S. Department of Energy (2008). 20% wind energy by 2030: Increasing
wind energys contribution to U.S. electricity supply. (DOE/GO-102008-
2567)
[36] Worldwatch Institute (2008). Renewables 2007: Global Status Report.
Washington, DC. : Martinot , E.
Wulff, H. E., (1966). The traditional cr
technology and influence on eastern and western civilization.MIT press.
284-289